Our previous work has provided evidence for extended major histocompatibility complex (MHC) haplotypes on caucasian chromosomes --specific alleles of HLA-B, DR, BF, C2, C4A, and C4B that occur as fixed sets with limited variation of alleles at other closely linked loci. There are at least 11 common extended haplotypes in whites and it is these, rather than any individual MHC alleles, that appear to "mark" a large number of diseases, including type I diabetes. If a disease susceptibility gene occurs on such a chromosome, all its alleles will be increased among patients. If it does not, its alleles will appear to be protective. Our proposed studies will futher characterize extended haplotypes in type I diabetics and use them to attempt to localize a susceptibility gene for this disease and to explore explanations for the excess of HLA-DR3/DR4 heterozygotes (over DR3 or DR4 homozygotes). Extended haplotypes will be studied in patients with Graves' disease, an if, as we think likely, there is a great increase in the extended haplotype (HLA-B8, DR3, SCO1) to explain the increases in HLA-B8 and DR3 obsered previously in this disease, this haplotype will be studied in detail including HLA-DP typing and restriction fragment length polymorphisms in Class II and Class III MHC genes and compared with that increased in diabetics. Detailed studies of the MHC In families with multiple type I diabetic sibs will be used to provide direct evidence for crossovers, for susceptibility of heterozygotes and for extra susceptibility genes in producing non-HLA-identifical affected sibs. Extended haploytypes as markers for age of onset, for complications and for immunological and other laboratory abnormalities in type I diabetes will be tested. Finally, the transmission of extended haplotypes from mothers and fathers will be analyzed in all study families to identify haplotypes with segregation distortion (possible human analogs of mouse t- mutants).